Identification of New Resistance Loci Against Sheath Blight Disease in Rice Through Genome-Wide Association Study

Chen Zongxiang, Feng Zhiming, Kang Houxiang , Zhao JianhuaChen TianxiaoLi QianqianGong Hongbing, Zhang YafangChen XijunPan XuebiaoLiu Wende Wang Guoliang Zuo Shimin

?

Identification of New Resistance Loci Against Sheath Blight Disease in Rice Through Genome-Wide Association Study

Chen Zongxiang1,#, Feng Zhiming1, #, Kang Houxiang2,#, Zhao Jianhua1, Chen Tianxiao1, Li Qianqian1, Gong Hongbing3, Zhang Yafang1, Chen Xijun1, Pan Xuebiao1, Liu Wende2, Wang Guoliang2, Zuo Shimin1, 2

(; State Key Laboratory for Biology of Plant Diseases and Insect Pests / Institute of Plant Protection, Chinese Academy of Agricultural Sciences, Beijing 100093, China; 212400, China; These authors contribute equally to this work)

Sheath blight (SB) caused by the soil borne pathogenis one of the most serious global rice diseases. Breeding resistant cultivar is the most economical and effective strategy to control the disease. However, no rice varieties are completely resistant to SB, and only a few reliable quantitative trait loci (QTLs) linked with SB resistance have been identified to date. In this study, we conducted a genome-wide association study (GWAS) of SB resistance using 299 varieties from the rice diversity panel 1 (RDP1) that were genotyped using 44 000 high-density single nucleotide polymorphism (SNP) markers. Through artificial inoculation, we found that only 36.5% of the tested varieties displayed resistance or moderate resistance to SB. In particular, theandsub-populations displayed higher SB resistance than the(TRJ),andsub-populations. Seven varieties showed similar resistance levels to the resistant control YSBR1. GWAS identified at least 11 SNP loci significantly associated with SB resistance in the three independent trials, leading to the identification of two reliable QTLs,and, on chromosomes 3 and 6. Using favorable alleles or haplotypes of significantly associated SNP loci, we estimated that both QTLs had obvious effects on reducing SB disease severity and can be used for enhancing SB resistance, especially in improving SB resistance of TRJ sub-population rice varieties. These results provided important information and genetic materials for developing SB resistant varieties through breeding.

genome-wide association study; quantitative trait locus; rice; sheath blight; plant height

Sheath blight (SB) is caused by the soil borne fungusKühn (), which poses a great threat to the rice grain yield and quality as one of the most serious diseases of rice (L.) worldwide (Lee and Rush, 1983; Marchetti and Bollich, 1991). Breeding resistant varieties is believed to be the most economical and effective strategy to control the disease compared to pesticide application. However, no rice germplasm has been identified with completely resistance to SB, and only partially resistance was reported. So far, only a few varieties showing stable resistance were reported, such as YSBR1, Tetep, Teqing and Jasmine 85 (Li et al, 1995; Chen et al, 2000; Li et al, 2000; Meena et al, 2000; Pinson et al, 2008; Jia et al, 2012).

SB resistance in rice is reported to be a typical quantitative trait, which is controlled by quantitative trait loci (QTLs) or multiple genes (Li et al, 1995; Zeng et al, 2010; Srinivasachary et al, 2011; Taguchi- Shiobara et al, 2013; Zuo et al, 2013, 2014; Eizenga et al, 2015; Chen et al, 2017; Jiang et al, 2018). Mapping QTLs for rice SB resistance has been increasingly emphasized. Li et al (1995) identified SB resistance QTLs for the first time using restricted fragment length polymorphism (RFLP) markers. Since then, more than 50 QTLs for SB resistance distributed on the rice 12 chromosomes have been detected using various mapping populations, such as F2populations (Pan et al, 1999a; Rush, 1999; Zou et al, 2000; Che et al, 2003; Arun et al, 2009), recombinant inbred lines (RILs) (Han et al, 2002; Pinson et al, 2005; Liu et al, 2009; Channamallikarjuna et al, 2010), chromosomal segment substitution lines (CSSLs) (Zuo et al, 2013, 2014), near-isogenic introgression lines (NILs) (Loan et al, 2004), double-haploid populations (DHs) (Kunihiro et al, 2002) and backcross populations (Sato et al, 2004; Tan et al, 2005; Eizenga et al, 2015). In addition, Jia L M et al (2012) conducted an association mapping with 155 simple sequence repeat (SSR) markers using 217 rice core germplasm, and identified 10 markers significantly associated with SB resistance. Sun et al (2014) detected 13 markers significantly associated with SB resistance through association analysis with 144 SSR markers using 456 rice varieties. However, only a few of these QTLs are real QTLs for SB resistance because they need to meet the criteria of being repeatedly detected in multiple environments and/or mapping populations and not being co-located with QTLs for plant height or heading date (Han et al, 2003; Sato et al, 2004; Pinson et al, 2005; Tan et al, 2005; Zuo et al, 2010, 2013, 2014;Wang et al, 2012; Eizenga et al, 2015). Moreover, only two QTLs for SB resistance (qSB-11andqSB-9) are fine-mapped while the remaining QTLs are only preliminarily mapped. No QTLs for SB resistance has been cloned to date (Zuo et al, 2013, 2014).

Accurate phenotyping is vital for mapping and further cloning of QTLs for complex traits such as SB resistance. The phenotype of SB resistance is influenced by many factors including plant height, heading date, planting density, temperature, humidity and soil fertility (Pinson et al, 2005; Jia et al, 2009). Controlling greenhouse or growth chamber conditions has been widely incorporated in the evaluation of SB resistant levels because controlled environmental conditions minimize the impact of other factors and provide more reliable data (Jia et al, 2007; Wamishe et al, 2007; Prasad and Eizenga, 2008; Liu et al, 2009; Jia L M et al, 2012; Sun et al, 2014), although most of the QTLs for SB resistance are identified in the field conditions.

The traditional genetic linkage method used to identify QTLs/genes is very time-consuming because it requires a large bi-parental mapping population and genotyping. More recently, genome-wide association study (GWAS), as a powerful approach, has been widely used to dissect a much broader genetic variability for complex traits in plants (Huang et al, 2010; Zhao et al, 2011; Morris et al, 2013; Kang et al, 2016; Liu et al, 2017). Compared to the traditional mapping method, GWAS generally employs more diverse natural populations and high-density single nucleotide polymorphism (SNP) markers, which helps in identifying marker loci more close to the candidate genes as well as in exploring favorable alleles of agronomic traits among natural varieties (Huang et al, 2010; Brachi et al, 2011; Zhao et al, 2011). In rice, many QTLs/genes related to grain quality, agronomic performance, biotic and abiotic stress have been characterized with GWAS (Huang et al, 2010; Famoso et al, 2011; Zhao et al, 2011; Kang et al, 2016; Zhu et al, 2016). However, very few studies identified SB resistance loci by the GWAS approach have been reported so far.

The rice diversity panel 1 (RDP1), which consists of approximately 420 phenotypically and genotypically diversevarieties collected from 82 countries, is divided into five major sub-populations [(TRJ),(TEJ),(IND),(AUS) and(ARO)] and an additional sub-population [admixture (ADM)] (Zhao et al, 2011; Eizenga et al, 2014; Zhu et al, 2016). Importantly, the RDP1 is genotyped with about 44 000 high-quality SNP markers (McCouch et al, 2010; Tung et al, 2010), and many QTLs/genes associated with various traits in rice are identified (Famoso et al, 2011; Zhao et al, 2011; Norton et al, 2014; Ueda et al, 2015; Kang et al, 2016).

In this study, we evaluated 299 RDP1 varieties for SB resistance and then conducted GWAS for the identification of SB resistance loci. As a result, several varieties with high SB resistance levels were found, and a number of SNP loci significantly associated with SB resistance were detected, which allowed us to identify two reliable QTLs for SB resistance. Results in this study provided critical information to further identification of SB resistance genes as well as developing SB resistant varieties through both marker-assisted selection and genomic selection.

Materials and Methods

Rice materials

In total, 299varieties, provided by the Genetic Stocks-(GSOR) Collection, USDA ARS Dale Bumpers National Rice Research Center, USA, were screened for SB resistance and used in GWAS. They represented the six sub-populations, including TRJ (66 varieties), TEJ (74 varieties), IND (58 varieties), AUS (44 varieties), ARO (11 varieties) and ADM (46 varieties) (Zhao et al, 2011; Supplemental Table 1). Five rice varieties with known SB resistance levels, Lemont (high susceptible), Wuyujing 3 (susceptible), Jasmine 85 (moderately resistant), C418 (moderately resistant) and YSBR1 (resistant) were served as references in the evaluation (Pan et al, 1999a; Zou et al, 2000; Chen et al, 2009; Liu et al, 2009; Zuo et al, 2009).

Sheath blight fungal inoculation

In order to create an ideal inoculation environment, the ‘mist-chamber’ was constructed in the greenhouse as reported previously (Wang et al, 2009; Supplemental Fig. 1-B). It was determined that the condition with temperature at 26 oC–30 oC and humidity at 75%–90% favored the growth ofon the plants in the mist-chamber.

In the greenhouse, the seedling-breeding plates (55 cm in length, 29 cm in width and 12 cm in depth) with 32 holes were filled with pre-sterilized soil. Five seeds were sown in each hole and thinned to four uniform seedlings at the third leaf stage were used for each variety for three replications. When the seedlings grew to the fourth leaf stage under the natural light condition, the plates were moved into the pre-built ‘mist-chamber’ and allowed 24 h for adaptation before pathogen inoculation.

The inoculation ofwas performed according to the method described previously (Zou et al, 2000; Jia et al, 2007; Zuo et al, 2013) with slight modification. The YN-7 strain (originally named RH-9) ofwith strong pathogenicity, provided by the Department of Plant Protection of Yangzhou University, China, was used for SB inoculation. Truncated thin matchsticks (0.8–1.0 cm in length, 2–3 mm in width and 1 mm in thickness) colonized by the YN-7 strain on the potato dextrose broth medium for 2–3 d at 28oC in the dark were used as the inocula (Supplemental Fig. 1-A). Each seedling in a hole was individually inoculated with an inoculum. Each inoculum was closely affixed to one side of the base of the seedling stem, assuring the hypha tightly touching the plant (Supplemental Fig. 1-C). The greenhouse condition was set for light for 13 h at 28 oC–30oC and dark for 11 h at 26oC–28oC, respectively.

Three independent trials of SB fungal inoculation described above were performed in August, June and September 2013, respectively.

Sheath blight symptom scoring

When the disease symptom appeared on the whole stems or leaves of susceptible control Lemont plant (generally at about 7–8 d after inoculation), the films of the ‘mist-chamber’ were immediately removed to rapidly reduce the inside humidity, which limits disease expansion. The SB disease scores of each seedling, calculated as the lesion length divided by the collar height (the length from the ground to the tallest leaf collar of the main stem) (Jia et al, 2007; Wang et al, 2009). Based on the ‘0–9’ rating score (RS), the RDP1 varieties were classified into six different resistant levels described previously: highly resistant (HR, RS ≤ 1.50), resistant (R, 1.50 < RS ≤ 3.00), moderately resistant (MR, 3.00 < RS ≤ 4.50), moderately susceptible (MS, 4.50 < RS ≤ 6.00), susceptible (S, 6.00 < RS ≤ 7.50) and highly susceptible (HS, 7.50 < RS ≤ 9.00) (Wang et al, 2009; Wang et al, 2011).

Data were processed with Microsoft Excel 2010. IBM SPSS version 16.0 (IBM Corp., Armonk, USA) was used to perform ANOVA and the Dunnett’smulti-comparison tests of the SB scores among different varieties or sub-populations.

GWAS analysis

The identification of SNP markers significantly associated with SB resistance by GWAS was performed according to Kang et al (2016). GWAS analysis was based on the publicly available 44 000-SNP dataset of RDP1 varieties (Zhao et al, 2011). The TASSEL 3.0 software and the mixed linear model (MLM) were used in GWAS (Bradbury et al, 2007). The MLM uses a joint kinship matrix and population structure model that can be described in Henderson’s matrix notation (Henderson, 1975). To control type I error, regions that had more than two SNPs with< 1 × 10-4within a 200-kb genomic window were considered for subsequent analysis. The Manhattan maps were plotted with PerL (Christiansen et al, 2012). EMMAX was used to fit a standard linear mixed model (Kang et al, 2010). Three principal component covariates were added to the model. Thepackage (https://cran. rproject.org/web/packages/qqman/) was used to produce Manhattan and quantile-quantile (QQ) plots.

Results

Less than 10% of the RDP1 varieties showed resistant reaction to R. solani with rating score lower than 3

A total of 299 RDP1 varieties and 5 reference varieties with known SB resistant levels were evaluated for SB resistance in greenhouse (Supplemental Table 1). Significant differences of SB resistance were found among varieties but not among three replicates within one trial (Supplemental Tables 2 and 3), indicating the high reproducibility of the evaluation test. Based on the average SB disease scores of each variety, we classified all the tested varieties into six different resistant reaction levels (Fig. 1). The five control varieties presented significant differences on SB disease scores (Supplemental Table 4), and were easily classified into four resistant levels, resistant YSBR1 (disease score was 2.86), moderately resistant C418 (3.84) and Jasmine 85 (4.26), susceptible Wuyujing 3 (5.49) and highly susceptible Lemont (8.29) (Fig. 1). This classification on the referent varieties is consistent with previous results (Chen et al, 2009; Wang et al, 2009; Zuo et al, 2009; Wang et al, 2011), indicating the reliability of the evaluation method.

The frequency distribution of varieties in different resistant levels showed that the majority of the 299 RDP1 varieties were moderately resistant or moderately susceptible to SB in all the three independent trials (Fig. 1). According to the average disease score, there were no completely immune or highly resistant varieties, and 3.7% (11 of 299) of the varieties were grouped as resistant and 32.8% (98 of 299) as moderately resistant (Fig. 1). These results indicated that the overall resistant level of the RDP1 varieties to SB was low, and only less than 10% showed resistant reaction with disease score lower than 3.

ARO and AUS were significantly more resistant than TRJ, IND and TEJ sub-populations

To understand the differences in SB resistance among sub-populations of the RDP1 varieties, we divided all the varieties into six sub-populations by structure analysis as reported previously (Zhu et al, 2016). We found that ARO and AUS sub-populations were significantly more resistant than the TRJ, IND and TEJ sub-populations (Table 1). The average disease scores of ARO and TRJ sub-populations were 4.09 and 5.45, respectively, which represented the lowest and the highest disease scores among sub-populations (Table 1).

Fig.1. Grouping of rice diversity panel 1 varieties in different reaction categories.

HR, Highly resistant; R, Resistant; MR, Moderately resistant; MS, Moderately susceptible; S, Susceptible; HS, Highly susceptible.YSBR1, C418, Jasmine 85, Wuyujing 3 and Lemont are five control varieties as R, MR, MS, S and HS, respectively.

Table 1.Multi-comparison of sheath blast disease scores of different sub-populations in the rice diversity panel 1.

TRJ,; IND,; TEJ,; ADM, Admixture; AUS,; ARO,.Values are Mean ± SD (n = 3). Different lowercase letters indicate significant difference at the 0.05 level.

Table 2.Number of varieties and the ratio in different types of resistant reaction in each sub-population.

TRJ,; TEJ,; IND,; ADM, Admixture; ARO,; AUS,; HR, Highly resistant; R, Resistant; MR, Moderately resistant; MS, Moderately susceptible; S, Susceptible; HS, Highly susceptible; RS, Rating score.

The ratio of varieties in different types of resistant reaction in each sub-populations was further analyzed (Table 2). We found that the sub-population with the highest ratio of resistant-type varieties was ARO (18.2%) and the lowest was IND (1.7%). On the contrary, for highly susceptible-type varieties, the highest ratio was IND (6.9%), and the lowest was ARO (0%) (Table 2). The ratio of varieties with resistant (R) and moderately resistant (MR) reactions in each sub-population was ranked in the descending order as follows: AUS (61.3%), ARO (54.5%), ADM (39.2%), IND (37.9%), TEJ (28.4%) and TRJ (20.7%) (Table 2).

SB resistance in the newly identified varieties was not due to their plant height

SB resistance is greatly influenced by morphological traits like plant height (Pinson et al, 2005; Jia et al, 2009). There were significant differences on seedling heights among the 299 rice germplasms (Supplemental Table 1), and linear regression analysis showed that the seedling height was negatively correlated with the disease score (= -0.231,< 0.001; Supplemental Fig. 2). To ensure that the resistant level of the newly screened resistant varieties was not due to the indirect effect from plant height, the plant height of the moderate resistant variety Jasmine 85 and the resistant variety YSBR1 were used as the references. We then compared the disease scores and plant heights between 11 resistant-type varieties (Table 3) and the five control varieties, and found that 7 (PR304, Ghati Kamma Nangarhar, Koshihikari, Bombilla, T26, Vary Vato 462 and Bico Branco) out of the 11 resistant-type varieties showed either lower or similar plant height compared with Jasmine 85 (Table 3). Among them, six showed similar SB resistance as the resistant control YSBR1, and one (Bico Branco) showed significantly higher SB resistance, indicating its high application potential in rice breeding against SB.

Two new QTLs for SB resistance were identified on chromosomes 3 and 6

Based on the 44 000 SNP data set and the disease scores of the RDP1 varieties, we identified 147 SNP markers significantly associated with SB resistance in trial 1, 21 in trial 2, and 11 in trial 3 (Fig. 2; Supplemental Table 5). More SNP markers in trial 1 maybe due to a larger phenotypic variation when scoring (Fig. 2-D). The contribution of each significant association marker on phenotypic variance was between 4.5% and 7.7% (Supplemental Table 5). These significant association markers were mainly distributed on rice chromosomes 1, 2, 3, 5, 6, 7, 8, 10 and 11 (Fig. 2). Among them, three SNP markers (id3008187, id6010523 and id6010787) were repeatedly detected in all the three trials, and seven SNP markers (ud5000337, id3008284, id6010496, id6011670, ud6000977, id6011721 and id7005052) were detected in two of the three trials. These 10 SNP markers were mainly located on chromosomes 3 and 6. Through the comparison of physical positions, we found that the two SNP markers (id3008187 and id3008284) on chromosome 3 were very near each other within 1 Mb apart. On chromosome 6, the six SNP markers (id6010523, id6010787, id6010496, id6011670, ud6000977 and id6011721) were mainly centralized in the 3 Mb region. Although we cannot determine the candidate resistance genes in the two regions, we can infer that each region contains a QTL for SB resistance, which is designated asand, respectively.

Table 3.Multi-comparisons of sheath blast disease score and plant height of the 5 control varieties and 11 new germplasms classified in the resistant type.

YSBR1, C418, Jasmine 85, Wuyujing 3 and Lemont are five control varieties as resistant, moderately resistant, moderately susceptible, susceptible, highly susceptible, respectively.ase scores and plant height..Values are Mean ± SD (= 3), and different lowercase letters indicate significant difference at the 0.05 level. Values are Mean ± SD (n = 3), and different lowercase letters indicate significant difference at the 0.05 level.

Favorable SNP haplotypes associated with QTLs qSB-3 and qSB-6 displaying obviously additive effects on SB resistance

For estimating the contribution of each QTL on SB resistance, we employed the most significant association SNP markers repeatedly detected in all the three trials in each QTL region to distinguish varieties with and without the resistant alleles. For, its most significant SNP locus was id3008187, in which ‘A’ was the favorable allele and ‘T’ was the unfavorable one. In total, 243 varieties contained ‘A’ (named A-type) and 42 varieties carried ‘T’ (named T-type). We found that in all the trials, the average disease score of A-type varieties was significantly lower than that of the T-type varieties, indicating the reliable contribution ofto SB resistance (Fig. 3-A). For, its most significant SNP loci were id6010523 and id6010787, in which the favorable haplotype is ‘GC’ (GC-haplotype) and the unfavorable is ‘TA’ (TA-haplotype). A total of 141 varieties contained the resistant GC-haplotype and 86 varieties harbored the susceptible TA-haplotype. The average disease score of GC-haplotype varieties was significantly lower than that of TA-haplotype varieties (Fig. 3-B). Moreover, the average disease score of AGC-haplotype varieties (128 varieties) that carried favorable alleles in bothandloci was even more significantly lower than that of TTA-haplotype varieties (38 varieties) carrying both unfavorable alleles (Fig. 3-C). According to the difference of average disease scores, we estimated the resistant effects of,and&to be 0.83, 0.70 and 1.21, respectively, in reducing the disease score (Fig. 3), suggesting an additive effect of the two QTLs. These results suggest that bothandhave obvious effects on SB resistance and an apparently pyramiding effect.

Fig. 2.Genome-wide association scan for single nucleotide polymorphism (SNP) loci associated with sheath blast (SB) resistance using rice diversity panel 1 accessions in three independent trials.

A–C, Manhattan plots of SNPs associated with SB resistance on 12 rice chromosomes in trial 1 (A), trial 2 (B) and trial 3 (C). The genomic coordinates are displayed along the-axis and the logarithm of the odds (LOD) score for each SNP is displayed on the-axis. The LOD score of each dot represents a transformedvalue, -lg(). Black horizontal lines indicate the genome-wide significance threshold. The arrow indicates the region containing at least two marker loci associated with SB resistance. The two regions marked by the rectangle boxes indicate the region repeatedly detected in the threetrials. D–F, Quantile-quantile (Q-Q) plots for the genome-wide association results to trial 1 (D), trial 2 (E) and trial 3 (F).

Further, most of varieties in the AUS (43/44) and ARO (9/11) sub-populations, but a few in the TEJ (20/74) and only one in TRJ (1/66) sub-populations, harbored the favorable haplotypes ‘AGC’ (Fig. 3-D), which is consistent with our above finding that ARO and AUS sub-populations were significantly more resistant than the TRJ and TEJ sub-populations (Table 1). In addition, interestingly, all the seven resistant- type varieties identified above belonged to the AGC- haplotype. In future, these associational SNP markers in theandregions can be applied to rice breeding programs to improve SB resistance through the marker-assisted selection strategy.

Discussion

Understanding the genetic architecture of resistance to rice SB is challenging because of the lack of reliable disease evaluation methods and appropriate mapping populations. In this study, we evaluated the SB resistance of 299 RDP1 varieties in three independent trials, and found that the overall resistant level of the RDP1 varieties was lower, and less than 10% of these varieties were ranked as the resistant (Fig. 1). Through GWAS with high density SNP markers, we identified 147 SNP loci significantly associated with SB resistance, and two reliable associated chromosomal regions (on chromosomes 3 and 6) that were repeatedly detected in the three trials (Fig. 2). The QTLs in these two regions were designated asand, and the favorable haplotypes of them were able to reduce SB scores by about 0.83 and 0.70, respectively (Fig. 3). We integrated the QTLs for SB resistance mapped so far on chromosomes 3 and 6 onto the physical map of the rice reference Nipponbare genome by their associated or flanking markers (Li et al, 1995; Pan et al, 1999b; Kunihiro et al, 2002; Sato et al, 2004; Pinson et al, 2005; Arun et al, 2009; Liu et al, 2009; Channamallikarjuna et al, 2010; Taguchi-Shiobara et al, 2013; Wang, 2013; Eizenga et al, 2015) (Fig. 4). From the integrated map, we found thatwas located in the same chromosomal region as a previously detected QTL (Wang, 2013), which is located in the physical interval of 13.3–18.2 Mb between markers RM3297 and RM3180 (Fig. 4-A). The location region ofdid not overlap with any previously detected QTLs on chromosome 6 (Fig. 4-B), implying thatis a newly discovered SB resistance locus. No previous reports are available that use GWAS with high density SNP markers to identify loci for SB resistance. Related linkage analysis populations through hybridization and backcrossing strategies will be constructed to further verify,and other association regions, as well as their disease-resistant effects. In addition, we are searching for potential candidate genes in theandregions and validating their functions in SB resistance using the CRISPR-Cas9 technology.

Fig. 3. Resistance effects and favorable haplotypes distribution ofand.

A, Effect of the favorable allele ‘A’ of id3008187 (represent) compared to the unfavorable allele ‘T’. B, Effect of the favorable haplotype ‘GC’ formed by id6010523 and id6010787 (represent) compared to the unfavorable haplotype ‘TA’. C, Effect of the favorable haplotype ‘AGC’ formed by id3008187, id6010523 and id6010787 compared to the unfavorable haplotype ‘TTA’. Trial 1, Trial 2, Trial 3 represent the three independent trials, respectively. Mean represents the average value of the three independent trials. **,< 0.01 by the Student’s-test. D, The ratio of favorable haplotypes (AGC-type) in each sub-population. TRJ, tropical japonica; TEJ, temperature japonica; IND, indica; ADM, Admixture; ARO, aromatic; AUS, aus.

Fig. 4. Integrated physical map of the QTLs for sheath blast resistance mapped so far on rice chromosomes 3 and 6.

The bars with different colors show the estimated location of these QTLs according to their flanking markers on the reference Nipponbare genome.

Two methods are available for the identification of rice SB resistance at the seedling stage, namely the ‘micro-chamber’ method (Jia et al, 2007) and the ‘mist-chamber’ method (Wang et al, 2009). The ‘mist-chamber’ method uses a large space and can accommodate more experimental materials at the same time compared to the ‘micro-chamber’ method, which helps improve the consistency of the testing environment. To further improve the efficiency and accuracy of scoring the SB disease when using the ‘mist-chamber’ method, the truncated thin matchsticks instead of rice husks served as thehypha carrier (Supplemental Fig. 1-A) because the use of matchsticks not only took a shorter time (about 3–4 d) to cultivate the pathogen but also ensured that the amount of hypha attached to the matchsticks was relatively consistent. Furthermore, obvious pathogen hypha were found on the base of the inoculated plants at 16 h after inoculation (Supplemental Fig. 1-C) and obvious disease lesions were visible within 2 d (data not shown). After 7 to 9 d, disease lesions expanded to the tallest leaf collar and even led to complete plant death for the highly susceptible cultivar Lemont (Supplemental Fig. 1-D). These results indicate that the matchsticks inoculum is suitable in the ‘mist- chamber’ environment.

The identification of SB resistance at the seedling stage can eliminate the influence of the different plant type and growth process among various varieties (Jia et al, 2007; Wang et al, 2009). Our study found that seedling height and disease score were negatively correlated (Supplemental Fig. 2), implying that the seedling heights of varieties may affect the development of SB disease. In order to exclude the influence of seedling height and identify truly resistant germplasms, the disease scores and seedling heights of the resistant variety YSBR1 and the moderately resistant variety Jasmine 85 were used as references. At last, seven varieties resistant to SB were identified. Among them, six showed similar SB resistance as the resistant control YSBR1, and one (Bico Branco) showed significantly higher SB resistance than YSBR1 similar (Table 3). The identification of these new disease-resistant germplasm has laid a valuable knowledge foundation and provides materials for breeding resistant varieties and mining resistance genes against SB.

The greatest advantage of GWAS is that it is likely to determine whether all the tested varieties carry favorable or unfavorable alleles in associated loci (Huang et al, 2010; Brachi et al, 2011; Zhao et al, 2011), which is unachievable by previous linkage analyses based on double parents or a few parents. Therefore, the data of GWAS can be more effectively combined with the practice of breeding. Using the information of the SNP marker loci associated with SB resistance, we can select suitable parents according to needs, and achieve rapid transfer of the target loci by carrying out marker-assisted selection and genomic breeding (Li et al, 2018). In addition, since the SB resistance is controlled by multiple genes, it is not practical to improve the overall resistance level of rice to SB only by marker-assisted polygenic polymerization. Therefore, it is still necessary to consciously increase the utilization frequency of resistant resources to SB in traditional breeding. In this study, we found that the ARO and AUS sub-populations were significantly more resistant than the TRJ, IND and TEJ sub-populations (Table 1), which is probably because most of varieties in AUS and ARO sub-populations carried the favorable haplotype ‘AGC’ inandregions (Fig. 3-D). According to the grouping result of the RDP1 varieties by Zhao et al (2011), most of the rice varieties in China belong to the IND and TEJ sub-populations. Therefore, in the future breeding of rice SB resistance in China, in addition to using known resistance sources, we can appropriately increase the application frequency of varieties in the AUS and ARO sub-populations, so as to generally enhance SB resistance of rice, especially in TRJ sub-population.

Acknowledgements

This work was partially supported by the Open Funding from State Key Laboratory for Biology of Plant Diseases and Insect Pests (Grant No. SKLOF201403), and by the Natural Science Foundation of China (Grant Nos. 31571748 and 31701057) and the Natural Science Foundation of Jiangsu Province, China (Grant Nos. BK20171293 and BK20141291), respectively.

SUPPLEMENTAL DATA

The following materials are available in the online versionof this article at http://www.sciencedirect.com/science/ journal/16726308; http://www.ricescience.org.

Supplemental Table 1. Germplasm used in assay for sheath blast resistance.

Supplemental Table 2. ANOVA of sheath blast disease scores of the five control varieties.

Supplemental Table 3. ANOVA of sheath blast disease scores of the tested RDP1 varieties.

Supplemental Table 4. Multi-comparison of sheath blast disease scores of the five control varieties.

Supplemental Table5. Markers significantly associated with sheath blast resistance.

Supplemental Fig. 1. Inoculum and inoculation in the mist-chamber assay and disease symptoms after inoculation.

Supplemental Fig. 2. Correlation analysis between sheath blast disease scores and plant heights of the RDP1 varieties.

Arun S, Annam M C, Shannonrm P, Josephl K, Arobert S, Rodantee T, Robert F. 2009. Genetic mapping of sheath blight resistance QTLs within tropicalrice cultivars., 49(1): 256–264.

Brachi B, Morris G P, Borevitz J O. 2011. Genome-wide association studies in plants: The missing heritability is in the field., 12(10): 232.

Bradbury P J, Zhang Z, Kroon D E, Casstevens T M, Ramdoss Y, Buckler E S. 2007. TASSEL: Software for association mapping of complex traits in diverse samples., 23(19): 2633–2635.

Channamallikarjuna V, Sonah H, Prasad M, Rao G J N, Chand S, Upreti H C, Singh N K, Sharma T R. 2010. Identification of major quantitative trait locifor sheath blight resistance in rice., 25(1): 155–166.

Che K P, Zhan Q C, Xing Q H, Wang Z P, Jin D M, He D J, Wang B. 2003. Tagging and mapping of rice sheath blight resistant gene., 106(2): 293–297.

Chen X J, Wang L, Zuo S M, Wang Z B, Chen Z X, Zhang Y F, Lu G D, Zhou E X, Guo Z J, Huang S W, Pan X B. 2009. Screening of varieties and isolates for identifying interaction between host and pathogen of rice sheath blight., 39: 514–520. (in Chinese with English abstract)

Chen Z X, Zou J H, Xu J Y, Tong Y H, Tang S Z, Wang Z B, Jiang R M, Ling B, Tang J, Pan X B. 2000. A preliminary study on resources of resistance to rice sheath blight., 14(1): 15–18. (in Chinese with English abstract)

Chen Z X, Feng Z M, Wang L P, Feng F, Zhang Y F, Ma Y Y, Pan X B, Zuo S M. 2017. Breeding potential of ricegene for tiller angle., 31(6): 590–598. (in Chinese with English abstract)

Christiansen T, Foy B D, Wall L, Orwant J. 2012. Programming Perl: Unmatched power for textprocessing and scripting.Covers Version 5.14. Sebastopol, CA, USA: O’Reilly.

Eizenga G C, Ali M L, Bryant R J, Yeater K M, McClung A M, McCouch S R. 2014. Registration of the rice diversity panel 1 for genome-wide association studies., 8(1): 109.

Eizenga G C, Jia M H, Pinson S R, Gasore E R, Prasad B. 2015. Exploring sheath blight quantitative trait loci in a Lemont/advanced backcross population., 35(6): 1–19.

Famoso A N, Zhao K, Clark R T, Tung C W, Wright M H, Bustamante C, Kochian L V, McCouch S R. 2011. Genetic architecture of aluminum tolerance in rice () determined through genome-wide association analysis and QTL mapping., 7(8): e1002221.

Han Y P, Xing Y Z, Chen Z X, Gu S L, Pan X B, Chen X L, Zhang Q F. 2002. Mapping QTLs for horizontal resistance to sheath blight in an elite rice restorer line, Minghui 63., 29(7): 622–626. (in Chinese with English abstract)

Han Y P, Xing Y Z, Gu S L, Chen Z X, Pan X B, Chen X L. 2003. Effect of morphological traits on sheath blight resistance in rice., 45(7): 825–831. (in Chinese with English abstract)

Henderson C R. 1975. Best linear unbiased estimation and prediction under a selection model., 31(2): 423–447.

Huang X H, Wei X H, Sang T, Zhao Q, Feng Q, Zhao Y, Li C Y, Zhu C R, Lu T T, Zhang Z W, Li M, Fan D L, Guo Y L, Wang A H, Wang L, Deng L W, Li W J, Lu Y Q, Weng Q J, Liu K Y, Huang T, Zhou T Y, Jing Y F, Li W, Lin Z, Buckler E S, Qian Q, Zhang Q F, Li J Y, Han B. 2010. Genome-wide association studies of 14 agronomic traits in rice landraces., 42: 961–967.

Jia L M, Yan W G, Zhu C S, Agrama H A, Jackson A, Yeater K, Li X, Huang B H, Hu B L, McClung A. 2012. Allelic analysis of sheath blight resistance with association mapping in rice., 7(3): e32703.

Jia Y, Correa-Victoria F, McClung A, Zhu L, Liu G, Wamishe Y, Xie J, Marchetti M A, Pinson S R M, Rutger J N. 2007. Rapid determination of rice cultivar responses to the sheath blight pathogenusing a micro-chamber screening method., 91(5): 485–489.

Jia Y, Liu G, Correa-Victoria F J, McClung A M, Oard J H, Bryant R J, Jia M H, Correll J C. 2012. Registration of four rice germplasm lines with improved resistance to sheath blight and blast diseases., 6(1): 95–100.

Jia Y L, Liu G J, Costanzo S, Lee S, Dai Y T. 2009. Current progress on genetic interactions of rice with rice blast and sheath blight fungi., 3(3): 231–239.

Jiang S F, Wang C J Z, Shu C W, Zhou E X. 2018. Cloning and expression analysis ofgene inAG-1IA of rice sheath blight pathogen., 32(2): 111–118.(in Chinese with English abstract)

Kang H M, Sul J H, Service S K, Zaitlen N A, Kong S, Freimer N B, Sabatti C, Eskin E. 2010. Variance component model to account for sample structure in genome-wide association studies., 42(4): 348–354.

Kang H X, Wang Y, Peng S S, Zhang Y L, Xiao Y H, Wang D, Qu S H, Li Z Q, Yan S Y, Wang Z L, Liu W D, Ning Y S, Korniliev P, Leung H, Mezey J, McCouch S R, Wang G L. 2016. Dissection of the genetic architecture of rice resistance to the blast fungus., 17(6): 959–972.

Kunihiro Y, Qian Q, Sato H, Teng S, Zeng D L, Fujimoto K, Zhu L H. 2002. QTL analysis of sheath blight resistance in rice (L.)., 29(1): 50–55. (in Chinese with English abstract)

Lee F N, Rush M C. 1983. Rice sheath blight: A major rice disease., 67(7): 829–832.

Li H, Song C Y, Cong W B, Wang G L. 2000. Evaluation and screening of resistance in keng rice varieties to sheath blight., 26(1): 19–21. (in Chinese with English abstract)

Li Y, Xiao J H, Chen L L, Huang X H, Cheng Z K, Han B, Zhang Q F, Wu C Y. 2018. Rice functional genomics research: Past decade and future., 11(3): 359–380.

Li Z K, Pinson S R M, Marchetti M A, Stansel J W, Park W D. 1995. Characterization of quantitative trait loci (QTLs) in cultivated rice contributing to field resistance to sheath blight ()., 91: 382–388.

Liu G, Jia Y, Correa-Victoria F J, Prado G A, Yeater K M, Mcclung A, Correll J C. 2009. Mapping quantitative trait loci responsible for resistance to sheath blight in rice., 99(9): 1078–1084.

Liu W Z, Maccaferri M, Chen X m, Laghetti G, Pignone D, Pumphrey M, Tuberosa R. 2017. Genome-wide association mapping reveals a rich genetic architecture of stripe rust resistance loci in emmer wheat (ssp.)., 130(11): 2249–2270.

Loan L C, Du P V, Li Z K. 2004. Molecular dissection of quantitative resistance of sheath blight in rice (L.).,12: 1–12.

Marchetti M A, Bollich C N. 1991. Quantification of the relationship between sheath blight severity and yield loss in rice., 75(8): 773.

McCouch S R, Zhao K Y, Wright M, Tung C W, Ebana K, Thomson M, Reynolds A, Wang D, Declerck G, Ali M L, McClung A, Eizenga G, Bustamante C. 2010. Development of genome-wide SNP assays for rice., 60(5): 524–535.

Meena B, Ramamoorthy V, Banu J G, Thangavelu R, Muthusamy M. 2000. Screening of rice genotypes against sheath blight disease., 12(2): 103–109.

Morris G P, Ramu P, Deshpande S P, Hash C T, Shah T, Upadhyaya H D, Riera-Lizarazu O, Brown P J, Acharya C B, Mitchell S E, Harriman J, Glaubitz J C, Buckler E S, Kresovich S. 2013. Population genomic and genome-wide association studies of agroclimatic traits in sorghum., 110(2): 453–458.

Norton G J, Douglas A, Lahner B, Yakubova E, Guerinot M L, Pinson S R M, Tarpley L, Eizenga G C, Mcgrath S P, Zhao F J, Islam S, Duan G, Zhu Y G, Salt D E, Meharg A A, Price A H. 2014. Genome wide association mapping of grain arsenic, copper, molybdenum and zinc in rice (L.) grown at four international field sites., 9(2): e89685.

Pan X B, Rush M C, Sha X Y, Xie Q J, Linscombe S D, Stetina S R, Oard J H. 1999a. Major gene, nonallelic sheath blight resistance from the rice cultivars Jasmine 85 and Teqing., 39(2): 338–346.

Pan X B, Zou J H, Chen Z X, Lu J F, Yu H X, Li H T, Wang Z B, Rush M C, Li Z. 1999b. Molecular marker localization of the main QTLs of rice variety Jasmine 85 resistant to sheath blight., 44: 1629–1635. (in Chinese with English abstract)

Pan X B, Zou J H, Chen Z X, Lu J F, Yu H X, Li H T, Wang Z B, Pan X Y, Rush M C, Zhu L H. 1999c. Tagging major quantitative trait loci for sheath blight resistance in a rice variety, Jasmine 85., 44(19): 1783–1789. (in Chinese with English abstract)

Pinson S R M, Capdevielle F M, Oard J H. 2005. Confirming QTLs and finding additional loci conditioning sheath blight resistance in rice using recombinant inbred lines., 45(2): 503–510.

Pinson S R M, Oard J H, Groth D, Miller R, Marchetti M A, Shank A R, Jia M H, Jia Y, Fjellstrom R G, Li Z. 2008. Registration of TIL:455, TIL:514, and TIL:642, three rice germplasm lines containing introgressed sheath blight resistance alleles., 2(3): 251–254.

Prasad B, Eizenga G C. 2008. Rice sheath blight disease resistance identified insppvarieties., 92(11): 1503–1509.

Sato H, Ideta O, Ando I, Kunihiro Y, Hirabayashi H, Iwano M, Miyasaka A, Nemoto H, Imbe T. 2004. Mapping QTLs for sheath blight resistance in the rice line WSS2., 54(3): 265–271.

Rush M C. 1999. Tagging major quantitative trait loci for sheath blight resistance in a rice variety, Jasmine 85., 44: 1783–1789. (in Chinese with English abstract).

Srinivasachary, Willocquet L, Savary S. 2011. Resistance to rice sheath blight (Kühn) [(teleomorph: Thanatephorus cucumeris (A.B. Frank) Donk.] disease: Current status and perspectives., 178(1): 1–22.

Sun X T, Lu D D, Ou-yang L J, Hu L F, Bian J M, Peng X S, Chen X R, Fu J R, He X P, He H H, Zhu C L. 2014. Association mapping and resistant alleles analysis for sheath blight resistance in rice., 40(5): 779. (in Chinese with English abstract)

Taguchi-Shiobara F, Ozaki H, Sato H, Maeda H, Kojima Y, Ebitani T, Yano M. 2013. Mapping and validation of QTLs for rice sheath blight resistance., 63(3): 301–308.

Tan C X, Ji X M, Yang Y, Pan X Y, Zuo S M, Zhang Y F, Zou J H, Chen Z X, Zhu L H, Pan X B. 2005. Identification and marker- assisted selection of two major quantitative genes controlling rice sheath blight resistance in backcross generations., 32(4): 399–405. (in Chinese with English abstract)

Tung C W, Zhao K, Wright M H, Ali M L, Jung J, Kimball J, Tyagi W, Thomson M J, Mcnally K, Leung H, Kim H Ahn S N, Reynolds A, Scheffler B, Eizenga G, McClung A, Bustamante C, McCouch A R. 2010. Development of a research platform for dissecting phenotype-genotype associations in rice (spp.)., 3(4): 205–217.

Ueda Y, Frimpong F, Qi Y, Matthus E, Wu L, H?ller S, Kraska T, Frei M. 2015. Genetic dissection of ozone tolerance in rice (L.) by a genome-wide association study., 66(1): 293–306.

Wamishe Y A, Jia Y, Singh P, Cartwright R D. 2007. Identification of field isolates ofto detect quantitative resistance in rice under greenhouse conditions., 1(4): 361–367.

Wang L, Huang W W, Liu L M, Fu Q, Huang S W. 2011. Evaluation of resistance to sheath blight () in somehybrid rice from southern China., 37(2): 263–270. (in Chinese with English abstract)

Wang Y, Prison S R M, Fjellstrom R G, Tabien R E. 2012. Phenotypic gain from introgression of two QTL,and, for rice sheath blight resistance., 30(1): 293–303.

Wang Z B, Zuo S M, Li G, Chen X J, Chen Z X, Zhang Y F, Pan X B. 2009. Rapid identification technology of resistance to rice sheath blight in seedling stage., 39(2): 174–182. (in Chinese with English abstract)

Wang Z B. 2013. New exploration of inoculation identification system for rice sheath blight and analysis of resistance to sheath blight in new germplasm YSBR1 [PhD Thesis]. Yangzhou: Yangzhou University. (in Chinese)

Zeng Y X, Li X M, Ma L Y, Ji Z J, Yang C D. 2010. Research progress on mapping of gene conferring resistance to sheath blight and exploitation of resistance resources in rice., 24(5): 544–550. (in Chinese with English abstract)

Zhao K, Tung C W, Eizenga G C, Wright M H, Ali M L, Price A H, Norton G J, Islam M R, Reynolds A, Mezey J, McClung A M, Bustamante C D, McCouch S R. 2011. Genome-wide association mapping reveals a rich genetic architecture of complex traits in., 2: 467.

Zhu D, Kang H, Li Z, Liu M, Zhu X, Wang Y, Wang D, Wang Z, Liu W, Wang G L. 2016. A genome-wide association study of field resistance toin rice., 9(1): 44.

Zou J H, Pan X B, Chen Z X, Xu J Y, Lu J F, Zhai W X, Zhu L H. 2000. Mapping quantitative trait loci controlling sheath blight resistance in two rice cultivars (L.)., 101: 569–573.

Zuo S M, Wang Z B, Chen X J, Gu F, Zhan Y F, Chen Z X, Pan X B, Pan C H. 2009, Evaluation of resistance of a novel rice germplasm YSBR1 to sheath blight., 35(4): 608–614. (in Chinese with English abstract)

Zuo S M, Zhang Y F, Chen Z X, Chen X J, Pan X B. 2010. Current progress in genetics and breeding in resistance to rice sheath blight., 40: 1014–1023. (in Chinese with English abstract)

Zuo S M, Yin Y J, Pan C H, Chen Z X, Zhang Y F, Gu S L, Zhu L H, Pan X B. 2013. Fine mapping ofqSB-11, the QTL that confers partial resistance to rice sheath blight., 126(5): 1257–1272.

Zuo S M, Zhang Y F, Yin Y J, Li G Z, Zhang G W, Wang H, Chen Z X, Pan X B. 2014. Fine-mapping ofqSB-9, a gene conferring major quantitative resistance to rice sheath blight., 34(4): 2191–2203.

20 August 2018;

29 October 2018

Zuo Shimin (smzuo@yzu.edu.cn); Wang Guoliang (wang.620@osu.edu)

Copyright ? 2019, China National Rice Research Institute. Hosting by Elsevier B V

This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Peer review under responsibility of China National Rice Research Institute

http://dx.doi.org/10.1016/j.rsci.2018.12.002

(Managing Editor: Wang Caihong)